Artificial Intelligence Models as a Guide to Conduct Bibliometric Analysis of Research on STEAM Learning

Abstract

The landscape of artificial intelligence is rapidly evolving, driven by the proliferation of sophisticated Large Language Models (LLMs). As demonstrated by Kasneci et al. (2023), models like ChatGPT are capable of generating educational content and providing on-demand support, highlighting their potential to transform learning environments. This ability to synthesize and generate information is fundamentally changing how knowledge is accessed and utilized. Furthermore, the multimodal capabilities of models like Gemini, which can process and integrate information from various sources (text, images, code), are expanding the scope of AI applications (Ruder, 2021). This study explores the transformative potential of DeepSeek's AI technologies in STEAM Learning (Science, Technology, Engineering, Arts, and Mathematics). The study presents a comprehensive bibliometric analysis of research at the intersection of STEAM (Science, Technology, Engineering, Arts, and Mathematics) learning over the past decade. Using data from repository; PubMed, researcher analyzed 37 publications to identify trends, key contributors, and emerging themes. The findings revealed a steady increase in research output since 2017, with significant contributions from the United States, China, and the United Kingdom. Leading journals such as the International Journal of STEM Education and Journal for Research in Mathematics Education dominate the field, while influential authors and institutions drive collaborative networks. This study provides a roadmap for future research, emphasizing the need for interdisciplinary collaboration and a focus on underrepresented sectors. The results provide insight for educators, policymakers, and researchers aiming to advance STEAM learning globally.